COVID-19 and driving behavior: Which were the most crucial influencing factors?

COVID-19 and driving behavior: Which were the most crucial influencing factors? | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article COVID-19 and driving behavior: Which were the most crucial influencing factors? Marios Sekadakis, Christos Katrakazas, Eva Michelaraki, Apostolos Ziakopoulos, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-2084342/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 08 Aug, 2023 Read the published version in Data Science for Transportation → Version 1 posted You are reading this latest preprint version Abstract This paper tries to identify and investigate the most significant factors that influenced the relationship between COVID-19 pandemic metrics (i.e., COVID-19 cases, fatalities and reproduction rate) and restrictions (i.e., stringency index and lockdown measures) with driving behavior in the entire 2020. To that aim, naturalistic driving data for a 12-month timeframe were exploited and analyzed. The examined driving behavior variables included harsh acceleration and harsh braking events concerning the time period before, during and after the lockdown measures in Greece. The harsh events were extracted using data obtained by a specially developed smartphone application which were transmitted to a back-end telematic platform between the 1st of January and the 31st of December, 2020. Based on the collected data, XGBoost feature analysis algorithms were deployed in order to obtain the most significant factors. Furthermore, a comparison among the first COVID-19 lockdown (i.e., February to May 2020), the second one (i.e., August to November 2020) and the period without COVID-19 restrictions was drawn. COVID-19 new cases and new fatalities were the most significant factors related to COVID-19 metrics impacting driving behavior. Additionally, the correlation between driving behavior with other factors (i.e., distance travelled, mobile use, driving requests, driving during risky hours) was revealed. Furthermore, the differences and similarities of the harsh events between the two lockdown periods were identified. This paper tries to fill this gap in existing literature concerning a feature analysis for the entire 2020 and including the first and second lockdown restrictions of the COVID-19 pandemic in Greece. COVID-19 pandemic Driving Behavior Harsh Brakings Harsh Accelerations Feature Analysis XGBoost Figures Figure 1 Figure 2 Figure 3 1. Introduction The COVID-19 pandemic has affected mobility patterns since December 2019 and continues incessantly for more than two years since the beginning (Zhu et al. 2020). Right from the beginning, many countries around the world imposed strict measures, such as lockdowns and suspension of all non-essential movements, in order to reduce human activity which contributes to the spread of the pandemic. In this direction, existing literature seeks to explore the dynamics of the pandemic in several countries around the world to understand the impact that COVID-19 had on the transport sector (Sharifi and Reza Khavarian-Garmsir 2020). As expected, the restriction measures affected typical patterns of travel activities and mobility in urban regions across the world (Kim 2021). It has been demonstrated that following the restrictive measures taken by governments to restrict the spread of the disease, an unprecedented decline in traffic volumes has been identified (Aletta et al. 2020; Katrakazas et al. 2020). For example, in the Netherlands, people reduced their outdoor activities due to the pandemic, leading to a decrease in the total number of trips and a reduction in distance travelled, with an increase in the proportion of people working from home (de Haas et al. 2020). Existing studies have also shown that there was a major change in the choice of transport mode, especially at the first pandemic wave, and consequently a change in the number of car-driven volumes was observed (Bucsky 2020). In the context of road safety, during the COVID-19 lockdown measures, the number of road collisions, injuries, and fatalities has significantly decreased, especially during the first lockdown period. This has been documented, for example, in particular, in the Spanish province of Tarragona, where a sharp decrease in traffic crashes was revealed (Saladié et al. 2020). Similarly, Carter (Carter 2020) showed that during the first COVID − 19 period (i.e., from March 15, 2020 to May 16, 2020), the total number of crashes in North Carolina decreased by half, fatalities decreased by 10%, and serious injuries increased by 6%, compared to the pre-closure baseline. A relevant study (Shilling and Waetjen 2020) indicated that all injury and fatal traffic crashes decreased on state highways and rural roads in California. Nevertheless, a study that used time-series to predict the road collisions, injuries and fatalities that would have been observed without the existence of the COVID-19 pandemic, made clear that the reduction of fatalities and injuries was disproportionate taking into account the reduction in traffic volumes (Sekadakis et al. 2021). Driving behavior has also changed during the pandemic as reported by recent studies (Katrakazas et al. 2020, 2021; Michelaraki et al. 2021). For example, according to the study by Katrakazas et al. (2020), which exploited driving data from the first lockdown period in Greece and Saudi Arabia, increased driving speed (6–11%) was observed, along with more frequent harsh accelerations and brakings per distance. Nevertheless, very few studies investigated driver behavior in more depth by analyzing and modeling naturalistic driving data. Katrakazas et al. (2021) quantified the impact of the pandemic COVID-19 on driving behavior using SARIMA time series modeling. The results showed that the observed values of three indicators of driving behavior (i.e., average speed, speeding, and harsh braking events per 100 km) were higher than the predicted values based on the corresponding observations before the first lockdown period in Greece. In this direction, the current study aims to identify and investigate the most significant factors in the entire 2020 that influenced the relationship between the COVID-19 pandemic metrics (i.e., COVID-19 cases, fatalities and reproduction rate) and restrictions (i.e., stringency index and lockdown measures) with driving behavior. For this purpose, naturalistic driving data for a 12-month timeframe were exploited and analyzed. The examined driving behavior variables were harsh acceleration and harsh braking events concerning a time period before, during and after the lockdown measures in Greece. The motivation is to cover the literature gap by giving insights on these two driving behavior indicators and how they influenced driving behavior for the entire year of 2020. A cross-lockdown comparison was also provided and gives insights into how the indicators varied across the examined conditions (i.e., no restrictions, 1st lockdown, 2nd lockdown). The paper structure is presented briefly: after the introduction, the methodology is described and includes the overview of the obtained dataset for this study, descriptive statistics of the examined variables, COVID-19 restriction measures and the chosen ML technique background are presented. Then, the analysis results are provided for both harsh acceleration events and harsh braking events. Finally, the main findings and conclusions are discussed, along with recommendations for further research. 2. Methodology 2.1 Data 2.1.1 Data Overview In order to correlate driving behavior with COVID-19 metrics and restrictions, OSeven Telematics (oseven.io) provided a random dataset with naturalistic driving trips from its database. The time span of the database was from 01/01/2020 to 31/12/2020 and included approximately 305,000 trips (randomly chosen) of trips throughout Greece. The aforementioned one-year dataset contains data before, during and after the first case of COVID-19 in Greece (i.e., 26/02/2020) and the imposition of two lockdowns for non-essential movements. OSeven exploits data from smartphone sensors (e.g. GPS, accelerometer data, and gyroscope data) using the smartphone applications and platform developed by OSeven Telematics. For each trip completed, a large amount of data was recorded, transmitted through Wi-Fi or cellular network and valuable critical information such as features, highlights and driving scores was produced in order to evaluate driving profile and performance. Subsequently, data were sent to the OSeven backend infrastructure, where there were evaluated using filtering, signal processing, ML algorithms and safety/eco scoring models. The OSeven platform has clear privacy policy statements and follows strict information security procedures, in compliance with the General Data Protection Regulation (GDPR) and related EU directives. Thus, all data has been provided by OSeven in a completely anonymized format and no geolocation information for the trips has been included in the dataset. Five variables (i.e. harsh accelerations (HA) /100km, harsh brakings (HB) /100km, mobile use/ driving time, driving during risky hours, distance) were exploited from the OSeven dataset and their description can be found in Table 1 . Furthermore, data from “Our World in Data” (OWD, 2020), were exploited in order to capture the daily evolution of COVID-19 metrics in 2020 i.e., new cases, new fatalities, and the COVID-19 reproduction rate of the pandemic. The response measures of the Greek government were quantified with the Stringency Index, by Oxford University and their COVID-19 government response tracker (Hale et al. 2020, 2021). Specifically, the stringency index ranges between 0 and 100 and represents the strictness of government responses to the pandemic. The stringency index is a composite measure based on 9 response indicators (i.e., school closing, workplace closing, cancel public events, restrictions on gatherings, close public transport, stay at home requirements, restrictions on internal movements, international travel controls, and public information campaigns) rescaled to a value from 0 to 100 (i.e., 100 = strictest response). In order to include traffic exposure data, the mobility data reports from Apple (Apple 2020) were used and specifically the driving requests as a surrogate measurement of traffic mobility. The aggregated data were collected from Apple Maps and show the mobility trends for major cities and several countries or regions. The information is generated by aggregating the number of daily driving requests made by the Apple Maps users who requested navigation. These requests are expressed by the percentage change compared to a baseline of 100% on January 13th, 2020, a date prior to the pandemic. All the driving variables examined in the current paper are summarized in Table 1 . Table 1 Variables Units, Description and Source Variable Unit Description Source Harsh accelerations (HA) /100km events/km Number of harsh accelerations per distance (100 km) OSeven Harsh brakings (HB) /100km events/km Number of harsh brakings per distance (100 km) OSeven Distance km Total trip distance OSeven Mobile Use/ Driving Time 0-100% Total duration of mobile usage in a trip/ Trip Duration OSeven Driving during Risky Hours km Distance driven in risky hours (00:00–05:00) in a trip OSeven New COVID-19 Cases count New confirmed cases of COVID-19 OWD New COVID-19 Fatalities count New fatalities attributed to COVID-19 OWD COVID-19 Reproduction Rate - Real-time estimate of the effective reproduction rate (R) of COVID-19 OWD Stringency Index 0-100 Government Response Stringency Index: composite measure based on 9 response indicators including school closures, workplace closures, and travel bans, rescaled to a value from 0 to 100 (100 = strictest response) Oxford Apple Driving Requests % change Requests for driving (%) (100% - baseline on January 13th, 2020) Apple Table 2 presents the descriptive statistics i.e., mean, standard deviation, maximum value, and minimum values of the investigated variables, for the random subset of trips (305,638 trips). More specifically, 16,927 trips (5.5% of the total) were observed during the 1st lockdown and 42,262 trips (13.8%) during the 2nd. It is worth noting that all the considered variables are continuous. The sample size was different for COVID-19 metrics, measures and mobility compared to driving data as they had daily observations for the entire 2020. The COVID-19 and mobility datasets derived from OWD, Oxford, and Apple were merged with each trip provided by OSeven into a mutual database for analysis purposes. Table 2 Descriptive Statistics of Investigated Variables Variable Mean SD Min Max Sample Size Harsh Accelerations (HA) /100km 9.36 17.37 0 99.98 305,638 Harsh Brakings (HB) /100km 13.59 19.76 0 99.99 305,638 Distance 13.28 23.65 0.50 648.69 305,638 Mobile Use/ Driving Time 0.05 0.14 0 1.00 305,638 Driving during Risky Hours 0.42 4.32 0 427.70 305,638 New COVID-19 Cases 363.40 662.56 0 3316.00 366 New COVID-19 Fatalities 12.36 26.93 0 121.00 366 COVID-19 Reproduction Rate 0.83 0.52 0 1.48 366 Stringency Index 48.17 28.11 0 84.26 366 Apple Driving Requests 114.35 52.57 18.59 241.14 364 SD: Standard Deviation 2.1.2 COVID-19 Restriction Measures Table 3 summarizes the two lockdown periods of non-essential movements due to the COVID-19 pandemic that have been announced by the Greek government. Table 3 Lockdown Measures and important Dates Greece – Lockdown Measures 1st Lockdown restrictions on non-essential movements 23-03-2020→04-05-2020 2nd Lockdown restrictions on non-essential movements 07-11-2020→31-12-2020 (Continued in 2021 ) The two lockdowns of 2020 are included in Fig. 1 in gray shades. Furthermore, the figure illustrates the evolution of driving mobility volumes (i.e., driving requests) through time in relation to COVID-19 new cases, stringency index of measures, and lockdown periods. An initial observation is that driving requests were significantly reduced during both lockdowns. Nevertheless, the greatest reduction in driving requests was during the first lockdown. Also, there was a spike of nearly 250% (150% more compared to the 100% baseline of January 13th ) in the requests in the first half of August. The illustrated restrictive periods along with the evolution of new COVID-19 cases, driving requests, and stringency index are expanded further into the analysis section. 2.2 XGBoost Analysis For analyzing harsh events per trip during the COVID-19 pandemic and extracting the most important factors, Extreme Gradient Boosting (XGBoost) algorithms were deployed. The XGBoost algorithms are supervised machine learning techniques that incorporate multiple Classification and Regression Trees (CART). In various ML competitions, XGBoost has regularly outperformed other approaches due to its versatility and efficiency (Nielsen 2016). XGBoost is used for supervised learning problems, where the training data (with multiple features) x i is used to predict a target variable y i (XGBoost Documentation 2021). Specifically, XGBoost algorithms were deployed in order to evaluate the feature importance of the aforementioned variables, i.e., driving requests, COVID-19 metrics and restrictions in regards to the naturalistic driving behavior indicators. The naturalistic driving behavior indicators were the frequency of harsh events, such as harsh brakings and harsh acceleration per distance (100km). It should be mentioned that XGBoost is a supervised ML technique and the user defines the independent/dependent variables. The learning process of the algorithm is iterative and consequently involves correcting previous errors in future iterations of the algorithm. A detailed overview of the comprised parameters and technical specifications of the algorithm in the used library can be found in (XGBoost Documentation 2021). XGBoost analysis was used because it has been shown to be superior in accuracy compared to logistic regression models or even other ML methods such as Random Forests, Artificial Neural Network, Support Vector Machines, both in the area of traffic safety (Huang and Meng 2019). In addition, the XGBoost algorithms have the capability to calculate the importance of each predictor variable in the developed model. In the XGBoost algorithm, the following three variable importance metrics were extracted (XGBoost Documentation 2021). These variable importance metrics are used by the XGBoost algorithms in the analysis to show which variables are informative in describing the driving behavior indicators (HA and HB /100km): Gain describes the enhancement in accuracy that a feature adds to its branches. Cover describes the relative amount of observations (or the number of samples) concerned by a feature. Frequency describes how often a feature is used in all generated trees. The gain metric is used for feature importance interpretation in the analysis. 2.3 XGBoost Parameters The XGBoost algorithm was run in an R-Studio environment using the xgboost package. Before running the algorithm, all the outliers were identified and then removed from the dataset, creating a clean undistorted set for analysis Then, a random split was employed in the data; 75% was the training set, while the remaining 25% was the test set. Furthermore, multiple values in terms of learning rate (eta) were tested (0.01–0.3) for each XGBoost for extracting the optimal model for harsh events. Learning rate is a tuning parameter in an optimization algorithm that determines the step size at each iteration while moving toward a minimum of a loss function (Murphy 2012). Additionally, K-fold cross validation was conducted in order to find the number of the best iteration within the XGBoost algorithm; preventing the model from overfitting. For each model, the function of K-fold cross validation tested about 200 different iterations in order to conclude the optimal iteration. The defined parameters for the XGBoost model for harsh events are provided as follows: Learning rate (eta) = 0.01–0.3 Gamma = 1 Maximum depth of a tree = 6 Subsample ratio of the training instances = 0.8 Subsample ratio of columns when constructing each tree = 0.5 2.4 Model Evaluation Model evaluation metrics were utilized to assess the predictive performance of the algorithm on the test set using the three metrics (i.e., Mean Error, Root Mean Squared Error, and Mean Absolute Error) indicated below, as a common practice. The \({e}_{t}\) represents the error, i.e., \({actual}_{t}-{predicted}_{t}\) , and N is the number of fitted points: Mean Error (ME): $$ME=\frac{1}{N} \sum _{i=1}^{N}{e}_{t} \left(1\right)$$ Root Mean Squared Error (RMSE): $$RMSE=\sqrt[2]{\frac{1}{N} \sum _{i=1}^{N}{{e}_{t}}^{2 }} \left(2\right)$$ Mean Absolute Error (MAE): $$MAE=\frac{1}{N} \sum _{i=1}^{N}\left|{e}_{t}\right| \left(3\right)$$ 3. Analysis And Results 3.1 Harsh Acceleration Events In this subsection, the results of the XGBoost model for Harsh Accelerations (HA) per 100km are presented. Firstly, the predictive power and accuracy provided by the application of the XGBoost algorithms on the test subset can be extracted by the achieved error. Table 4 presents the accomplished error i.e., ME = 0.081, RMSE = 17.314 and MAE = 12.012. Table 4 Errors on Test Predictions ME RMSE MAE Test set 0.081 17.314 12.012 The obtained feature importance is provided in Table 5 . The top three variables that impacted HA/100km were distance, mobile use/ driving time, and driving requests. Also, a small contribution was provided by driving during risky night-time hours. With regards to COVID-19 new cases in Greece seems to affect HA the most. The reproduction rate of the virus, as well as the strictness of measures and the new rates of fatalities due to COVID-19 have less impact on HA/100km. The detailed influence of predicting harsh events as expressed by the gain scores of XGBoost is shown in Table 5 . Table 5 Feature importance of HA/100km - XGBoost algorithms Feature Gain Cover Frequency Distance 0.531 0.364 0.242 Mobile use/ Driving time 0.207 0.212 0.198 Apple Driving Requests 0.086 0.174 0.161 New COVID-19 Cases 0.060 0.083 0.115 Driving during Risky hours 0.039 0.053 0.107 COVID-19 Reproduction Rate 0.031 0.042 0.078 Stringency Index 0.023 0.029 0.045 New COVID-19 Fatalities 0.023 0.043 0.054 Furthermore, boxplots were created supplementary to XGBoost in order to reveal the trend of harsh accelerations under the three aforementioned restriction measures of 2020 (i.e., 1st Lockdown, 2nd Lockdown and the time period without restrictions). These boxplots are given in Fig. 2 . The boxplot shows the median, interquartile range, minimum, and maximum values of completion time for each measure. Figure 2 (a), presents the boxplot for harsh accelerations including the whole dataset. As can be seen in the boxplot, the median values for each condition are equal to zero. On contrary to the non-zero mean value of harsh accelerations in Table 2 , zero median values are extracted due to the fact that the values are not normally distributed and many of the trips recorded zero frequency in terms of harsh acceleration per distance, and thus the lower quartile (25th percentile) is equal to median as well. Consequently, an additional boxplot in Fig. 2 (b) was created by excluding the zero values of the dataset. Hence, this boxplot presents only the trips with harsh events occurrence since trips with zero harsh acceleration frequency were excluded. As shown in Fig. 2 (b), the 2nd lockdown in Greece had a narrower interquartile range than the 1st lockdown and the period without restrictions. This means that the upper quartile (i.e., 75th percentile) of the 2nd lockdown is lower than the other conditions. Investigating the interquartile range of the boxplot the range of harsh events can be depicted and thus concluded the different patterns among different restrictive conditions. In this respect, the lower upper quartile of the 2nd lockdown reveals that the majority of the observed values (between 25th and 75th percentile) had lower values and range than the other restrictive conditions. Also, the 1st lockdown has a higher upper quartile compared to without restrictions and the 2nd lockdown. The highest median was observed at the 1st lockdown, then at 2nd, and then without restrictions. 3.2 Harsh Braking Events In this subsection, the results for Harsh Brakings (HB)/100km are presented and, as mentioned previously, these outcomes are further elaborated in the subsequent Discussion section. Table 6 presents the accomplished errors for this model i.e., ME=-0.025, RMSE = 19.529 and MAE = 14.561. Table 6 Errors on Test Predictions ME RMSE MAE Test set -0.025 19.529 14.561 The obtained feature importance is provided in Table 7 . Similar to the harsh accelerations model, the top three variables that impacted the most the HB were; distance, mobile use/ driving time and the number of driving requests by Apple. A small contribution was also provided by driving during risky night-time hours. However, the COVID-19-related variable that influenced the most HB in Greece was found to be different than the HA model. COVID-19 Reproduction Rate was found to influence the most HB. Other COVID-19-related variables that influenced the frequency of harsh brakings in Greece were new COVID-19 Cases, the Stringency Index and the number of new COVID-19 fatalities. Table 7 Feature importance of HB/100km - XGBoost algorithms Feature Gain Cover Frequency Distance 0.608 0.368 0.250 Mobile use/ Driving time 0.102 0.132 0.182 Apple Driving Requests 0.078 0.151 0.140 COVID-19 Reproduction Rate 0.063 0.082 0.085 New COVID-19 Cases 0.053 0.101 0.124 Driving during Risky hours 0.039 0.077 0.117 Stringency Index 0.039 0.053 0.051 New COVID-19 Fatalities 0.019 0.035 0.051 Similar to the model for harsh accelerations, Fig. 3 (a) presents the corresponding boxplot for harsh brakings including the entire dataset. As can be seen in this boxplot, the highest median value was observed during the 1st lockdown. Then, the conditions without restrictions follow and it is noteworthy that the median for the 2nd lockdown equals zero. The box of the 2nd lockdown in Greece has a narrower interquartile range than the other conditions (i.e., 1st lockdown and without restrictions). This means that the upper quartile of the 2nd lockdown is lower than the other conditions. In this respect, the lower upper quartile reveals that the majority of the observed harsh braking event values (between 25th and 75th percentile) had lower values and range than the other restrictive conditions. Also, the 1st lockdown has a higher upper quartile compared to without restrictions and the 2nd lockdown. Figure 3 (b), similarly to the HA model, the highest median was observed at the 1st lockdown, then at 2 nd, and then without restrictions. 4. Discussion The current paper aims to identify and investigate the most significant factors that influenced driving behavior during 2020, a year that behavior was heavily influenced by the COVID-19 pandemic. Both COVID-19 metrics (i.e., COVID − 19 cases, fatalities, and reproduction rate) and restrictions (i.e., stringency index and lockdown measures) were taken into account to identify their relationship with driving behavior. The XGBoost algorithm was chosen as the analysis method and the results suggest a strong correlation between COVID-19 metrics and restriction measures with driving behavior. Furthermore, different patterns were revealed for both harsh events among three examined conditions, i.e., without restrictions, 1st lockdown, and 2nd lockdown. Modelling results demonstrated that there are three common crucial factors that influenced HA and HB events the most during the pandemic. These factors were distance, mobile use/ driving time, and driving requests (requested in Apple Maps). More specifically, trip distance and mobile use duration were the two most important factors out of the eight examined variables that influence HA and HB. Τrip distance had a great impact on HA and HB events probably due to the fact that the longer trips were driven on highways and rural roads than trips within urban environment. Hence, the change in road type probably influences the drivers’ braking and acceleration patterns with more or less frequent harsh events. Another causal factor for the correlation between harsh events and duration was the increasing fatigue by increasing the trip distance. However, these assumptions need further research in order to be validated. Additionally, mobile phone use shows the importance of drivers being undistracted in order to avoid HA and HB events. After trip duration and mobile phone use, driving requests follow which are a driving exposure measurement and is an indication of the prevailing traffic volumes. This finding reveals the relation between this exposure measurement with HA and HB events. A higher value of exposure indicates a greater density of traffic and by extension, it changes the probabilities of the driver being involved in a harsh event for instance with more dense surrounding traffic. A small contribution to HA and HB was also provided by driving during risky nighttime hours indicating that there was a change in events during nighttime driving (00:00–05:00) due to the lighting conditions themselves as well as it was probably affected due to the prohibitions imposed by the Greek government during the nighttime and essentially reduced trips during risky hours (Katrakazas et al. 2021). The three aforementioned variables are evidently not directly related to the pandemic. Nevertheless, four COVID-19-related variables were found to impact HA and HB events. New COVID-19 cases in Greece were found to prevail compared to other COVID-19-related variables in terms of HA events. On contrary to HA, COVID-19 Reproduction Rate was found to influence HB events the most. The most influential pandemic-related factors for HA and HB events in Greece were COVID-19 Reproduction Rate, Stringency Index, and New COVID-19 Fatalities and Cases. This is in line with existing literature. For example, the studies of (Dong et al. 2022; Lee et al. 2020; Vanlaar et al. 2021) found that COVID-19 restrictions negatively affected risky driving behaviors such as speeding, and distracted driving. With regards to traffic exposure during 2020, it can be concluded from Fig. 1 that driving requests were significantly decreased during both lockdowns compared to the baseline of no restrictions. The greatest reduction was observed in the first lockdown compared to the second. This means that the traffic volume during the 1st lockdown was lower than in the other conditions (i.e., during the 2nd lockdown, and without restrictions). Hence, with fewer vehicles ahead, the drivers could accelerate more easily and this can be revealed in Fig. 2 (a), where the upper quartile was higher than in other conditions. Additionally, in Fig. 2 (b) for trips with harsh accelerations occurrence, the median was higher during the 1st lockdown than the other conditions (i.e., during 2nd lockdown, and without restrictions) meaning that the HA events were more frequent. This finding can be related partly to speeding, an increase was revealed in the spatial extent of speeding, and in the level of speeding as well as statistically significant differences in speeding before and after the COVID-19 outbreak (Lee et al. 2020). With regards to the 2nd lockdown, for trips with harsh accelerations, the median was higher compared to conditions without restrictions, as a result of the decreased traffic volume but not at the same magnitude as the 1st lockdown, in which the traffic volume was much lower. With regards to HB events, in Fig. 3 (a), again, the upper quartile is greater during the 1st lockdown than other conditions (i.e., during 2nd lockdown, and without restrictions) and combining Fig. 3 (b), the median is higher for trips with HB occurrence and this implies that the HB events were more frequent. This finding is also consistent with the literature (Katrakazas et al. 2020). This can be explained, for instance, as the traffic volume during the 1st lockdown was lower than the other conditions and hence with fewer vehicles ahead, the drivers could maintain higher speeds, as stated in (Katrakazas et al. 2020). With higher speeds, the drivers were more probable to be involved in a harsh braking event with potential traffic obstacles ahead (i.e., pedestrians, bikes, scooters and traffic control signs or signals), especially during the lockdowns that the active transport was increased (Linares-Rendón and Garrido-Cumbrera 2021). With regards to the 2nd lockdown following the same logic as HA, for trips with harsh brakings, the median was higher compared to no restrictions as a result of the decreased traffic volume but not the same magnitude as the 1st in which the traffic volume was lower. The results of the exploratory analysis by XGBoost indicate a correlation of COVID-19 metrics and restrictive measures with harsh brakings and accelerations. This correlation could be explained as COVID-19 metrics and restriction measures affected commuters by leading them to stay at home. Consequently, the stay at home restrictions led to a decreased traffic volume, this can be validated by Fig. 1 , and thus the traffic volumes affected directly driving behavior. This phenomenon can substantiate why the driving requests are a more important factor in the analysis than COVID-19-related variables. Nevertheless, this work is not without shortcomings, and therefore, future research could focus on covering the remaining gaps that this work did not cover. Initially, future studies could concentrate on more sophisticated models, such as deep neural networks, e.g., Convolutional neural networks (CNNs) or Artificial Neural Networks (ANNs), which probably can accomplish lower errors and give more insights into driving behavior variables. In addition, more variables with regards to driving behavior, i.e., speeding, speeding duration, and speed, could be exploited using the same method in order to give in the same context results. These variables were tested but they led to models with large errors and therefore, were not included in this work. Finally, additional data with geolocation and road type information could also enhance the current methodology, leading to spatial analyses of the examined variables. 5. Conclusions The present paper aims at identifying the most important factors that influenced driving behavior during 2020, a year that behavior was heavily influenced by the COVID-19 pandemic. In order to accomplish this study, naturalistic driving data for a 12-month timeframe were exploited and analyzed. The examined driving behavior variables were HA and HB events per 100km concerning the time period before, during and after the imposition of lockdown measures in Greece in 2020, the first year of the COVID-19 pandemic. The naturalistic driving data were extracted by a specially developed smartphone application and were transmitted to a back-end telematic platform from OSeven Telematics. The top three variables that influenced the most HA and HB events were distance, mobile use/ driving time, and driving requests (as requested in Apple Maps). Focusing on the COVID-19-related variables, this study identified that the most significant factors in the entire 2020 were new COVID-19 cases, new COVID-19 fatalities, COVID-19 reproduction rate as well as stringency index. The results of the exploratory analysis by XGBoost indicate a correlation of COVID-19 metrics and restriction measures with harsh brakings and accelerations. As mentioned previously, COVID-19 restrictions affected commuters to stay at home, and subsequently with lower traffic volumes; driving behavior was affected. Furthermore, for all the investigated three conditions, i.e., no restrictions, 1st lockdown, and 2nd lockdown, different HA and HB event patterns were revealed. HA and HB events were more frequent and with higher values range during the 1st and then 2nd lockdown compared to non-restrictive conditions for trips with harsh events occurrence due to their correlation with driving exposure measurements (i.e., Apple driving requests). Moreover, it was found that HA and HB events were also affected by risky nighttime hours, indicating that there was a change in events during nighttime driving (00:00–05:00) due to the lighting conditions themselves as well as probably affected due to the prohibitions imposed by the Greek government during the nighttime and essentially reduced trips during risky hours. Declarations Ethical Approval The ethics guidelines have been approved by the director of the Department of Transportation Planning and Engineering of the School of Civil Engineering at the National Technical University of Athens. The authors declare that the current study was conducted according to the ethical principles of the Declaration of Helsinki since no one was harmed physically or mentally affected during the driving measurements, and the drivers participated voluntarily. Competing interests On behalf of all authors, the corresponding author states that there is no conflict of interest. Authors' contributions Marios Sekadakis: Conceptualization, Methodology, Writing – original draft, Analysis, Data curation. Christos Katrakazas: Conceptualization, Methodology, Writing – review & editing. Eva Michelaraki: Writing – original draft, review & editing. Apostolos Ziakopoulos: Analysis, Data curation. George Yannis: Supervision, Resources. All authors reviewed the manuscript. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Availability of data and materials The datasets analysed during the current study are not publicly available due to are confidential in nature and may only be provided with restrictions (e.g. anonymized data) by OSeven Telematics, London, UK (www.oseven.io). Acknowledgment The authors would like to thank OSeven Telematics, London, UK (oseven.io) for providing all the necessary naturalistic driving data that assisted in accomplishing this study. References Aletta, F., Brinchi, S., Carrese, S., Gemma, A., Guattari, C., Mannini, L., and Patella, S. M. (2020). “Analysing urban traffic volumes and mapping noise emissions in Rome (Italy) in the context of containment measures for the COVID-19 disease.” Noise Mapping , 7(1), 114–122. Apple. (2020). “COVID‑19 - Mobility Trends Reports - Apple [WWW Document].” Accessed 11/06/20 , (Jun. 11, 2020). Bucsky, P. (2020). “Modal share changes due to COVID-19 : The case of Budapest.” Transportation Research Interdisciplinary Perspectives , The Author, 8, 100141. Carter, D. (2020). Effects of COVID-19 Shutdown on Crashes and Travel in NC . Dong, X., Xie, K., and Yang, H. (2022). “How did COVID-19 impact driving behaviors and crash Severity? A multigroup structural equation modeling.” Accident Analysis and Prevention , Elsevier Ltd, 172(December 2021), 106687. de Haas, M., Faber, R., and Hamersma, M. (2020). “How COVID-19 and the Dutch ‘intelligent lockdown’ change activities, work and travel behaviour: Evidence from longitudinal data in the Netherlands.” Transportation Research Interdisciplinary Perspectives , Elsevier Ltd, 6, 100150. Hale, T., Anania, J., Angrist, N., Boby, T., Cameron-Blake, E., Folco, M. Di, Ellen, L., Goldszmidt, R., Hallas, L., KIra, B., Luciano, M., Majumadar, S., Nagesh, R., Petherick, A., Phillips, T., Tatlow, H., Webster, S., Wood, A., and Zhang, Y. (2021). Variation in government responses to COVID-19, Version 12.0 . BSG Working Paper Series . Hale, T., Angrist, N., Cameron-Blake, E., Hallas, L., Kira, B., Majumdar, S., Petherick, A., Phillips, T., Tatlow, H., and Webster, S. (2020). “Variation in government responses to COVID-19.” BSG Working Paper Series. Blavatnik School of Government. University of Oxford , Version 8.0. Huang, Y., and Meng, S. (2019). “Automobile insurance classification ratemaking based on telematics driving data.” Decision Support Systems , Elsevier, 127(January 2019), 113156. Katrakazas, C., Michelaraki, E., Sekadakis, M., and Yannis, G. (2020). “A descriptive analysis of the effect of the COVID-19 pandemic on driving behavior and road safety.” Transportation Research Interdisciplinary Perspectives , Elsevier Ltd, 7, 100186. Katrakazas, C., Michelaraki, E., Sekadakis, M., Ziakopoulos, A., Kontaxi, A., and Yannis, G. (2021). “Identifying the impact of the COVID-19 pandemic on driving behavior using naturalistic driving data and time series forecasting.” Journal of Safety Research , National Safety Council and Elsevier Ltd, 78, 189–202. Kim, K. (2021). “Impacts of COVID-19 on transportation: Summary and synthesis of interdisciplinary research.” Transportation Research Interdisciplinary Perspectives , Volume 9(100305). Lee, J., Porr, A., and Miller, H. (2020). “Evidence of Increased Vehicle Speeding in Ohio ’ s Major Cities during the COVID-19 Pandemic.” 1–6. Linares-Rendón, F., and Garrido-Cumbrera, M. (2021). “Impact of the COVID-19 Pandemic on Urban Mobility: A Systematic Review of the Existing Literature.” Journal of Transport & Health . Michelaraki, E., Sekadakis, M., Katrakazas, C., Ziakopoulos, A., and Yannis, G. (2021). “A four-country comparative overview of the impact of COVID-19 on traffic safety behavior.” 10th International Congress on Transportation Research , Rhodes, Greece. Murphy, K. P. (2012). Machine learning : a probabilistic perspective . Nielsen, D. (2016). “Tree Boosting With XGBoost: Why Does XGBoost Win ‘Every’ Machine Learning Competition?” Department of Mathematical Sciences. Our World in Data. (2020). “Coronavirus Pandemic (COVID-19).” (Dec. 8, 2020). Saladié, Ò., Bustamante, E., and Gutiérrez, A. (2020). “Transportation Research Interdisciplinary Perspectives COVID-19 lockdown and reduction of traf fi c accidents in Tarragona province, Spain.” Transportation Research Interdisciplinary Perspectives , The Authors, 8, 100218. Sekadakis, M., Katrakazas, C., Michelaraki, E., Kehagia, F., and Yannis, G. (2021). “Analysis of the impact of COVID-19 on collisions, fatalities and injuries using time series forecasting: The case of Greece.” Accident Analysis & Prevention , Elsevier Ltd, 162(July), 106391. Sharifi, A., and Reza Khavarian-Garmsir, A. (2020). “The COVID-19 pandemic : Impacts on cities and major lessons for urban planning, design, and management.” Science of the Total Environment , 749, 1–3. Shilling, F., and Waetjen, D. (2020). “Impact of COVID19 mitigation on numbers and costs of California Traffic Crashes.” 2020. Vanlaar, W. G. M., Woods-Fry, H., Barrett, H., Lyon, C., Brown, S., Wicklund, C., and Robertson, R. D. (2021). “The impact of COVID-19 on road safety in Canada and the United States.” Accident Analysis and Prevention , Elsevier Ltd, 160, 106324. XGBoost Documentation. (2021). “dmlc XGBoost.” (Nov. 26, 2021). Zhu, N., Zhang, D., Wang, W., Li, X., Yang, B., Song, J., Zhao, X., Huang, B., Shi, W., Lu, R., Niu, P., Zhan, F., Ma, X., Wang, D., Xu, W., Wu, G., Gao, G. F., and Tan, W. (2020). “A Novel Coronavirus from Patients with Pneumonia in China, 2019.” New England Journal of Medicine , 382(8), 727–733. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 08 Aug, 2023 Read the published version in Data Science for Transportation → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-2084342","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":146200059,"identity":"29be91ca-4228-4606-8492-31bb0b9f04b5","order_by":0,"name":"Marios Sekadakis","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+0lEQVRIiWNgGAWjYHACZgYGNgYDNgbmAwwMBhYkaWFLAGqRAHGI1MLAwAPEDERo0e0/fNjgR5mdMR//ma8bfhRIMJjL9xh+YPh1D6cWsxtpyYk955LN2CRyt93sATrMso3HWIKxrxiPFh7jA7xtzDZsErzbbvAAtRgcA5KMPQm4tZw///ng37Z6Gzb+M89u/oFoMf6BV8uBHOZk3rbDZmwMOWy3obaYSTD8wKPlRpqxscy548ZsEmlmt2UMJHgMjqWVWSQ24HPY4ceSb8qqDef3H352880fGzmDw4c33/jwB7cWDMADJhPbiNcBA39I1zIKRsEoGAXDFgAAtLZPNOZV0nYAAAAASUVORK5CYII=","orcid":"","institution":"National Technical University of Athens","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Marios","middleName":"","lastName":"Sekadakis","suffix":""},{"id":146200060,"identity":"897c1d16-c219-41c8-809e-c749a4dbf406","order_by":1,"name":"Christos Katrakazas","email":"","orcid":"","institution":"National Technical University of Athens","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Christos","middleName":"","lastName":"Katrakazas","suffix":""},{"id":146200061,"identity":"220c0272-3f2e-458d-b686-5c9e01fee44d","order_by":2,"name":"Eva Michelaraki","email":"","orcid":"","institution":"National Technical University of Athens","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Eva","middleName":"","lastName":"Michelaraki","suffix":""},{"id":146200062,"identity":"5a9f525d-21f3-4877-9295-6b56c6e715cd","order_by":3,"name":"Apostolos Ziakopoulos","email":"","orcid":"","institution":"National Technical University of Athens","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Apostolos","middleName":"","lastName":"Ziakopoulos","suffix":""},{"id":146200063,"identity":"114823a1-11db-4acf-ae25-005634a95dc0","order_by":4,"name":"George Yannis","email":"","orcid":"","institution":"National Technical University of Athens","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"George","middleName":"","lastName":"Yannis","suffix":""}],"badges":[],"createdAt":"2022-09-20 10:14:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-2084342/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-2084342/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s42421-023-00078-7","type":"published","date":"2023-08-08T21:55:27+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":28245081,"identity":"3e898041-4559-4981-a478-ac49352f2055","added_by":"auto","created_at":"2022-10-25 19:01:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":168255,"visible":true,"origin":"","legend":"\u003cp\u003eOverview of mobility along with COVID-19 restrictions (lockdown and stringency index) and new cases\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-2084342/v1/209d1e54512ef0864e92853b.png"},{"id":28245080,"identity":"919e8202-440a-4ea5-9321-2ab684f0196f","added_by":"auto","created_at":"2022-10-25 19:01:36","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":135254,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Harsh Accelerations/100km under different restriction measures\u003c/p\u003e\n\u003cp\u003e(b) Harsh Accelerations/100km under different restriction measures by excluding zero values\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-2084342/v1/96affff3cd1a4172cf1c392f.png"},{"id":28245221,"identity":"d36397a7-c8c3-4aa4-b482-36b8145eff0b","added_by":"auto","created_at":"2022-10-25 19:06:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":144134,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Harsh Brakings/100km under different restriction measures\u003c/p\u003e\n\u003cp\u003e(b) Harsh Brakings/100km under different restriction measures by excluding zero values\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-2084342/v1/404154c58bdf090d1aafb5bd.png"},{"id":44736476,"identity":"549bf892-e5c6-44b7-8eb3-b0851a31ac54","added_by":"auto","created_at":"2023-10-16 22:30:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":788388,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-2084342/v1/7633c492-821f-4267-b1f8-fe42fa08b2e1.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"COVID-19 and driving behavior: Which were the most crucial influencing factors?","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe COVID-19 pandemic has affected mobility patterns since December 2019 and continues incessantly for more than two years since the beginning (Zhu et al. 2020). Right from the beginning, many countries around the world imposed strict measures, such as lockdowns and suspension of all non-essential movements, in order to reduce human activity which contributes to the spread of the pandemic.\u003c/p\u003e \u003cp\u003eIn this direction, existing literature seeks to explore the dynamics of the pandemic in several countries around the world to understand the impact that COVID-19 had on the transport sector (Sharifi and Reza Khavarian-Garmsir 2020). As expected, the restriction measures affected typical patterns of travel activities and mobility in urban regions across the world (Kim 2021). It has been demonstrated that following the restrictive measures taken by governments to restrict the spread of the disease, an unprecedented decline in traffic volumes has been identified (Aletta et al. 2020; Katrakazas et al. 2020). For example, in the Netherlands, people reduced their outdoor activities due to the pandemic, leading to a decrease in the total number of trips and a reduction in distance travelled, with an increase in the proportion of people working from home (de Haas et al. 2020). Existing studies have also shown that there was a major change in the choice of transport mode, especially at the first pandemic wave, and consequently a change in the number of car-driven volumes was observed (Bucsky 2020).\u003c/p\u003e \u003cp\u003eIn the context of road safety, during the COVID-19 lockdown measures, the number of road collisions, injuries, and fatalities has significantly decreased, especially during the first lockdown period. This has been documented, for example, in particular, in the Spanish province of Tarragona, where a sharp decrease in traffic crashes was revealed (Saladi\u0026eacute; et al. 2020). Similarly, Carter (Carter 2020) showed that during the first COVID \u0026minus;\u0026thinsp;19 period (i.e., from March 15, 2020 to May 16, 2020), the total number of crashes in North Carolina decreased by half, fatalities decreased by 10%, and serious injuries increased by 6%, compared to the pre-closure baseline. A relevant study (Shilling and Waetjen 2020) indicated that all injury and fatal traffic crashes decreased on state highways and rural roads in California. Nevertheless, a study that used time-series to predict the road collisions, injuries and fatalities that would have been observed without the existence of the COVID-19 pandemic, made clear that the reduction of fatalities and injuries was disproportionate taking into account the reduction in traffic volumes (Sekadakis et al. 2021).\u003c/p\u003e \u003cp\u003eDriving behavior has also changed during the pandemic as reported by recent studies (Katrakazas et al. 2020, 2021; Michelaraki et al. 2021). For example, according to the study by Katrakazas et al. (2020), which exploited driving data from the first lockdown period in Greece and Saudi Arabia, increased driving speed (6\u0026ndash;11%) was observed, along with more frequent harsh accelerations and brakings per distance. Nevertheless, very few studies investigated driver behavior in more depth by analyzing and modeling naturalistic driving data. Katrakazas et al. (2021) quantified the impact of the pandemic COVID-19 on driving behavior using SARIMA time series modeling. The results showed that the observed values of three indicators of driving behavior (i.e., average speed, speeding, and harsh braking events per 100 km) were higher than the predicted values based on the corresponding observations before the first lockdown period in Greece.\u003c/p\u003e \u003cp\u003eIn this direction, the current study aims to identify and investigate the most significant factors in the entire 2020 that influenced the relationship between the COVID-19 pandemic metrics (i.e., COVID-19 cases, fatalities and reproduction rate) and restrictions (i.e., stringency index and lockdown measures) with driving behavior. For this purpose, naturalistic driving data for a 12-month timeframe were exploited and analyzed. The examined driving behavior variables were harsh acceleration and harsh braking events concerning a time period before, during and after the lockdown measures in Greece. The motivation is to cover the literature gap by giving insights on these two driving behavior indicators and how they influenced driving behavior for the entire year of 2020. A cross-lockdown comparison was also provided and gives insights into how the indicators varied across the examined conditions (i.e., no restrictions, 1st lockdown, 2nd lockdown).\u003c/p\u003e \u003cp\u003eThe paper structure is presented briefly: after the introduction, the methodology is described and includes the overview of the obtained dataset for this study, descriptive statistics of the examined variables, COVID-19 restriction measures and the chosen ML technique background are presented. Then, the analysis results are provided for both harsh acceleration events and harsh braking events. Finally, the main findings and conclusions are discussed, along with recommendations for further research.\u003c/p\u003e"},{"header":"2. Methodology","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data\u003c/h2\u003e \u003cdiv id=\"Sec4\" class=\"Section3\"\u003e \u003ch2\u003e2.1.1 Data Overview\u003c/h2\u003e \u003cp\u003eIn order to correlate driving behavior with COVID-19 metrics and restrictions, OSeven Telematics (oseven.io) provided a random dataset with naturalistic driving trips from its database. The time span of the database was from 01/01/2020 to 31/12/2020 and included approximately 305,000 trips (randomly chosen) of trips throughout Greece. The aforementioned one-year dataset contains data before, during and after the first case of COVID-19 in Greece (i.e., 26/02/2020) and the imposition of two lockdowns for non-essential movements. OSeven exploits data from smartphone sensors (e.g. GPS, accelerometer data, and gyroscope data) using the smartphone applications and platform developed by OSeven Telematics. For each trip completed, a large amount of data was recorded, transmitted through Wi-Fi or cellular network and valuable critical information such as features, highlights and driving scores was produced in order to evaluate driving profile and performance. Subsequently, data were sent to the OSeven backend infrastructure, where there were evaluated using filtering, signal processing, ML algorithms and safety/eco scoring models. The OSeven platform has clear privacy policy statements and follows strict information security procedures, in compliance with the General Data Protection Regulation (GDPR) and related EU directives. Thus, all data has been provided by OSeven in a completely anonymized format and no geolocation information for the trips has been included in the dataset.\u003c/p\u003e \u003cp\u003eFive variables (i.e. harsh accelerations (HA) /100km, harsh brakings (HB) /100km, mobile use/ driving time, driving during risky hours, distance) were exploited from the OSeven dataset and their description can be found in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eFurthermore, data from \u0026ldquo;Our World in Data\u0026rdquo; (OWD, 2020), were exploited in order to capture the daily evolution of COVID-19 metrics in 2020 i.e., new cases, new fatalities, and the COVID-19 reproduction rate of the pandemic.\u003c/p\u003e \u003cp\u003eThe response measures of the Greek government were quantified with the Stringency Index, by Oxford University and their COVID-19 government response tracker (Hale et al. 2020, 2021). Specifically, the stringency index ranges between 0 and 100 and represents the strictness of government responses to the pandemic. The stringency index is a composite measure based on 9 response indicators (i.e., school closing, workplace closing, cancel public events, restrictions on gatherings, close public transport, stay at home requirements, restrictions on internal movements, international travel controls, and public information campaigns) rescaled to a value from 0 to 100 (i.e., 100\u0026thinsp;=\u0026thinsp;strictest response).\u003c/p\u003e \u003cp\u003eIn order to include traffic exposure data, the mobility data reports from Apple (Apple 2020) were used and specifically the driving requests as a surrogate measurement of traffic mobility. The aggregated data were collected from Apple Maps and show the mobility trends for major cities and several countries or regions. The information is generated by aggregating the number of daily driving requests made by the Apple Maps users who requested navigation. These requests are expressed by the percentage change compared to a baseline of 100% on January 13th, 2020, a date prior to the pandemic.\u003c/p\u003e \u003cp\u003eAll the driving variables examined in the current paper are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eVariables Units, Description and Source\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnit\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHarsh accelerations (HA) /100km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eevents/km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of harsh accelerations per distance (100 km)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOSeven\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHarsh brakings (HB) /100km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eevents/km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of harsh brakings per distance \u003c/p\u003e \u003cp\u003e(100 km)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOSeven\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDistance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ekm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTotal trip distance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOSeven\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMobile Use/ Driving Time\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0-100%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTotal duration of mobile usage in a trip/ \u003c/p\u003e \u003cp\u003eTrip Duration\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOSeven\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDriving during Risky Hours\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ekm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDistance driven in risky hours (00:00\u0026ndash;05:00) in a trip\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOSeven\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNew COVID-19 Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecount\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNew confirmed cases of COVID-19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOWD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNew COVID-19 Fatalities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ecount\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNew fatalities attributed to COVID-19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOWD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCOVID-19 Reproduction Rate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eReal-time estimate of the effective reproduction rate (R) of COVID-19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOWD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStringency Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0-100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGovernment Response Stringency Index: composite measure based on 9 response indicators including school closures, workplace closures, and travel bans, rescaled to a value from 0 to 100 \u003c/p\u003e \u003cp\u003e(100\u0026thinsp;=\u0026thinsp;strictest response)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOxford\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eApple Driving Requests\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e% change\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRequests for driving (%) \u003c/p\u003e \u003cp\u003e(100% - baseline on January 13th, 2020)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eApple\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the descriptive statistics i.e., mean, standard deviation, maximum value, and minimum values of the investigated variables, for the random subset of trips (305,638 trips). More specifically, 16,927 trips (5.5% of the total) were observed during the 1st lockdown and 42,262 trips (13.8%) during the 2nd. It is worth noting that all the considered variables are continuous. The sample size was different for COVID-19 metrics, measures and mobility compared to driving data as they had daily observations for the entire 2020. The COVID-19 and mobility datasets derived from OWD, Oxford, and Apple were merged with each trip provided by OSeven into a mutual database for analysis purposes.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive Statistics of Investigated Variables\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eMean\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eSD\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eMin\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eMax\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003eSample Size\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHarsh Accelerations (HA)\u003c/p\u003e \u003cp\u003e/100km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e17.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e99.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e305,638\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHarsh Brakings (HB) \u003c/p\u003e \u003cp\u003e/100km\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e19.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e99.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e305,638\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDistance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e648.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e305,638\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMobile Use/ Driving Time\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e305,638\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDriving during Risky Hours\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e427.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e305,638\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNew COVID-19 Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e363.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e662.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3316.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e366\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNew COVID-19 Fatalities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e12.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e26.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e121.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e366\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCOVID-19 Reproduction Rate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e366\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStringency Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e48.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e28.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e84.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e366\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eApple Driving Requests\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e114.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e52.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e241.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e364\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eSD: Standard Deviation\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.1.2 COVID-19 Restriction Measures\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarizes the two lockdown periods of non-essential movements due to the COVID-19 pandemic that have been announced by the Greek government.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eLockdown Measures and important Dates\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eGreece \u0026ndash; Lockdown Measures\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1st Lockdown restrictions on non-essential movements\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23-03-2020\u0026rarr;04-05-2020\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2nd Lockdown restrictions on non-essential movements\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e07-11-2020\u0026rarr;31-12-2020 (Continued in 2021 )\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe two lockdowns of 2020 are included in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e in gray shades. Furthermore, the figure illustrates the evolution of driving mobility volumes (i.e., driving requests) through time in relation to COVID-19 new cases, stringency index of measures, and lockdown periods. An initial observation is that driving requests were significantly reduced during both lockdowns. Nevertheless, the greatest reduction in driving requests was during the first lockdown. Also, there was a spike of nearly 250% (150% more compared to the 100% baseline of January 13th ) in the requests in the first half of August. The illustrated restrictive periods along with the evolution of new COVID-19 cases, driving requests, and stringency index are expanded further into the analysis section.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.2 XGBoost Analysis\u003c/h2\u003e \u003cp\u003eFor analyzing harsh events per trip during the COVID-19 pandemic and extracting the most important factors, Extreme Gradient Boosting (XGBoost) algorithms were deployed. The XGBoost algorithms are supervised machine learning techniques that incorporate multiple Classification and Regression Trees (CART). In various ML competitions, XGBoost has regularly outperformed other approaches due to its versatility and efficiency (Nielsen 2016). XGBoost is used for supervised learning problems, where the training data (with multiple features) x\u003csub\u003ei\u003c/sub\u003e is used to predict a target variable y\u003csub\u003ei\u003c/sub\u003e (XGBoost Documentation 2021). Specifically, XGBoost algorithms were deployed in order to evaluate the feature importance of the aforementioned variables, i.e., driving requests, COVID-19 metrics and restrictions in regards to the naturalistic driving behavior indicators. The naturalistic driving behavior indicators were the frequency of harsh events, such as harsh brakings and harsh acceleration per distance (100km). It should be mentioned that XGBoost is a supervised ML technique and the user defines the independent/dependent variables. The learning process of the algorithm is iterative and consequently involves correcting previous errors in future iterations of the algorithm. A detailed overview of the comprised parameters and technical specifications of the algorithm in the used library can be found in (XGBoost Documentation 2021). XGBoost analysis was used because it has been shown to be superior in accuracy compared to logistic regression models or even other ML methods such as Random Forests, Artificial Neural Network, Support Vector Machines, both in the area of traffic safety (Huang and Meng 2019).\u003c/p\u003e \u003cp\u003eIn addition, the XGBoost algorithms have the capability to calculate the importance of each predictor variable in the developed model. In the XGBoost algorithm, the following three variable importance metrics were extracted (XGBoost Documentation 2021). These variable importance metrics are used by the XGBoost algorithms in the analysis to show which variables are informative in describing the driving behavior indicators (HA and HB /100km):\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eGain describes the enhancement in accuracy that a feature adds to its branches.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eCover describes the relative amount of observations (or the number of samples) concerned by a feature.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFrequency describes how often a feature is used in all generated trees.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThe gain metric is used for feature importance interpretation in the analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.3 XGBoost Parameters\u003c/h2\u003e \u003cp\u003eThe XGBoost algorithm was run in an R-Studio environment using the \u003cem\u003exgboost\u003c/em\u003e package. Before running the algorithm, all the outliers were identified and then removed from the dataset, creating a clean undistorted set for analysis Then, a random split was employed in the data; 75% was the training set, while the remaining 25% was the test set. Furthermore, multiple values in terms of learning rate (eta) were tested (0.01\u0026ndash;0.3) for each XGBoost for extracting the optimal model for harsh events. Learning rate is a tuning parameter in an optimization algorithm that determines the step size at each iteration while moving toward a minimum of a loss function (Murphy 2012).\u003c/p\u003e \u003cp\u003eAdditionally, K-fold cross validation was conducted in order to find the number of the best iteration within the XGBoost algorithm; preventing the model from overfitting. For each model, the function of K-fold cross validation tested about 200 different iterations in order to conclude the optimal iteration.\u003c/p\u003e \u003cp\u003eThe defined parameters for the XGBoost model for harsh events are provided as follows:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLearning rate (eta)\u0026thinsp;=\u0026thinsp;0.01\u0026ndash;0.3\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eGamma\u0026thinsp;=\u0026thinsp;1\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eMaximum depth of a tree\u0026thinsp;=\u0026thinsp;6\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eSubsample ratio of the training instances\u0026thinsp;=\u0026thinsp;0.8\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eSubsample ratio of columns when constructing each tree\u0026thinsp;=\u0026thinsp;0.5\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Model Evaluation\u003c/h2\u003e \u003cp\u003eModel evaluation metrics were utilized to assess the predictive performance of the algorithm on the test set using the three metrics (i.e., Mean Error, Root Mean Squared Error, and Mean Absolute Error) indicated below, as a common practice. The \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({e}_{t}\\)\u003c/span\u003e\u003c/span\u003e represents the error, i.e., \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({actual}_{t}-{predicted}_{t}\\)\u003c/span\u003e\u003c/span\u003e, and \u003cem\u003eN\u003c/em\u003e is the number of fitted points:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eMean Error (ME):\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv id=\"Equa\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$ME=\\frac{1}{N} \\sum _{i=1}^{N}{e}_{t} \\left(1\\right)$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eRoot Mean Squared Error (RMSE):\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv id=\"Equb\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$RMSE=\\sqrt[2]{\\frac{1}{N} \\sum _{i=1}^{N}{{e}_{t}}^{2 }} \\left(2\\right)$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eMean Absolute Error (MAE):\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv id=\"Equc\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$MAE=\\frac{1}{N} \\sum _{i=1}^{N}\\left|{e}_{t}\\right| \\left(3\\right)$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Analysis And Results","content":"\u003cdiv class=\"Section2\" id=\"Sec10\"\u003e\n \u003ch2\u003e3.1 Harsh Acceleration Events\u003c/h2\u003e\n \u003cp\u003eIn this subsection, the results of the XGBoost model for Harsh Accelerations (HA) per 100km are presented. Firstly, the predictive power and accuracy provided by the application of the XGBoost algorithms on the test subset can be extracted by the achieved error. Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e presents the accomplished error i.e., ME\u0026thinsp;=\u0026thinsp;0.081, RMSE\u0026thinsp;=\u0026thinsp;17.314 and MAE\u0026thinsp;=\u0026thinsp;12.012.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable border=\"1\" id=\"Tab4\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eErrors on Test Predictions\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eME\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRMSE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMAE\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTest set\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.081\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17.314\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12.012\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eThe obtained feature importance is provided in Table \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e. The top three variables that impacted HA/100km were distance, mobile use/ driving time, and driving requests. Also, a small contribution was provided by driving during risky night-time hours. With regards to COVID-19 new cases in Greece seems to affect HA the most. The reproduction rate of the virus, as well as the strictness of measures and the new rates of fatalities due to COVID-19 have less impact on HA/100km. The detailed influence of predicting harsh events as expressed by the gain scores of XGBoost is shown in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable border=\"1\" id=\"Tab5\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eFeature importance of HA/100km - XGBoost algorithms\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFeature\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGain\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCover\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFrequency\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.531\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.364\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.242\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMobile use/ Driving time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.207\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.212\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.198\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eApple Driving Requests\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.086\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.174\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.161\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNew COVID-19 Cases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.060\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.083\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.115\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDriving during Risky hours\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.107\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCOVID-19 Reproduction Rate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.031\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.042\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.078\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStringency Index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.029\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.045\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNew COVID-19 Fatalities\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.054\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eFurthermore, boxplots were created supplementary to XGBoost in order to reveal the trend of harsh accelerations under the three aforementioned restriction measures of 2020 (i.e., 1st Lockdown, 2nd Lockdown and the time period without restrictions). These boxplots are given in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e. The boxplot shows the median, interquartile range, minimum, and maximum values of completion time for each measure. Figure \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e (a), presents the boxplot for harsh accelerations including the whole dataset. As can be seen in the boxplot, the median values for each condition are equal to zero. On contrary to the non-zero mean value of harsh accelerations in Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e, zero median values are extracted due to the fact that the values are not normally distributed and many of the trips recorded zero frequency in terms of harsh acceleration per distance, and thus the lower quartile (25th percentile) is equal to median as well. Consequently, an additional boxplot in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e (b) was created by excluding the zero values of the dataset. Hence, this boxplot presents only the trips with harsh events occurrence since trips with zero harsh acceleration frequency were excluded. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e (b), the 2nd lockdown in Greece had a narrower interquartile range than the 1st lockdown and the period without restrictions. This means that the upper quartile (i.e., 75th percentile) of the 2nd lockdown is lower than the other conditions. Investigating the interquartile range of the boxplot the range of harsh events can be depicted and thus concluded the different patterns among different restrictive conditions. In this respect, the lower upper quartile of the 2nd lockdown reveals that the majority of the observed values (between 25th and 75th percentile) had lower values and range than the other restrictive conditions. Also, the 1st lockdown has a higher upper quartile compared to without restrictions and the 2nd lockdown. The highest median was observed at the 1st lockdown, then at 2nd, and then without restrictions.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Section2\" id=\"Sec11\"\u003e\n \u003ch2\u003e3.2 Harsh Braking Events\u003c/h2\u003e\n \u003cp\u003eIn this subsection, the results for Harsh Brakings (HB)/100km are presented and, as mentioned previously, these outcomes are further elaborated in the subsequent \u003cspan class=\"InternalRef\"\u003eDiscussion\u003c/span\u003e section. Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e presents the accomplished errors for this model i.e., ME=-0.025, RMSE\u0026thinsp;=\u0026thinsp;19.529 and MAE\u0026thinsp;=\u0026thinsp;14.561.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable border=\"1\" id=\"Tab6\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eErrors on Test Predictions\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eME\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRMSE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMAE\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTest set\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.025\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19.529\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e14.561\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eThe obtained feature importance is provided in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e. Similar to the harsh accelerations model, the top three variables that impacted the most the HB were; distance, mobile use/ driving time and the number of driving requests by Apple. A small contribution was also provided by driving during risky night-time hours. However, the COVID-19-related variable that influenced the most HB in Greece was found to be different than the HA model. COVID-19 Reproduction Rate was found to influence the most HB. Other COVID-19-related variables that influenced the frequency of harsh brakings in Greece were new COVID-19 Cases, the Stringency Index and the number of new COVID-19 fatalities.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003ctable border=\"1\" id=\"Tab7\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eFeature importance of HB/100km - XGBoost algorithms\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFeature\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGain\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCover\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFrequency\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.608\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.368\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.250\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMobile use/ Driving time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.102\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.132\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.182\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eApple Driving Requests\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.078\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.151\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.140\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCOVID-19 Reproduction Rate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.063\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.082\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.085\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNew COVID-19 Cases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.101\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.124\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDriving during Risky hours\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.077\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.117\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStringency Index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNew COVID-19 Fatalities\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.035\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eSimilar to the model for harsh accelerations, Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e (a) presents the corresponding boxplot for harsh brakings including the entire dataset. As can be seen in this boxplot, the highest median value was observed during the 1st lockdown. Then, the conditions without restrictions follow and it is noteworthy that the median for the 2nd lockdown equals zero. The box of the 2nd lockdown in Greece has a narrower interquartile range than the other conditions (i.e., 1st lockdown and without restrictions). This means that the upper quartile of the 2nd lockdown is lower than the other conditions. In this respect, the lower upper quartile reveals that the majority of the observed harsh braking event values (between 25th and 75th percentile) had lower values and range than the other restrictive conditions. Also, the 1st lockdown has a higher upper quartile compared to without restrictions and the 2nd lockdown. Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e (b), similarly to the HA model, the highest median was observed at the 1st lockdown, then at 2\u003csup\u003end,\u003c/sup\u003e and then without restrictions.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe current paper aims to identify and investigate the most significant factors that influenced driving behavior during 2020, a year that behavior was heavily influenced by the COVID-19 pandemic. Both COVID-19 metrics (i.e., COVID \u0026minus;\u0026thinsp;19 cases, fatalities, and reproduction rate) and restrictions (i.e., stringency index and lockdown measures) were taken into account to identify their relationship with driving behavior. The XGBoost algorithm was chosen as the analysis method and the results suggest a strong correlation between COVID-19 metrics and restriction measures with driving behavior. Furthermore, different patterns were revealed for both harsh events among three examined conditions, i.e., without restrictions, 1st lockdown, and 2nd lockdown.\u003c/p\u003e \u003cp\u003eModelling results demonstrated that there are three common crucial factors that influenced HA and HB events the most during the pandemic. These factors were distance, mobile use/ driving time, and driving requests (requested in Apple Maps). More specifically, trip distance and mobile use duration were the two most important factors out of the eight examined variables that influence HA and HB. Τrip distance had a great impact on HA and HB events probably due to the fact that the longer trips were driven on highways and rural roads than trips within urban environment. Hence, the change in road type probably influences the drivers\u0026rsquo; braking and acceleration patterns with more or less frequent harsh events. Another causal factor for the correlation between harsh events and duration was the increasing fatigue by increasing the trip distance. However, these assumptions need further research in order to be validated. Additionally, mobile phone use shows the importance of drivers being undistracted in order to avoid HA and HB events. After trip duration and mobile phone use, driving requests follow which are a driving exposure measurement and is an indication of the prevailing traffic volumes. This finding reveals the relation between this exposure measurement with HA and HB events. A higher value of exposure indicates a greater density of traffic and by extension, it changes the probabilities of the driver being involved in a harsh event for instance with more dense surrounding traffic. A small contribution to HA and HB was also provided by driving during risky nighttime hours indicating that there was a change in events during nighttime driving (00:00\u0026ndash;05:00) due to the lighting conditions themselves as well as it was probably affected due to the prohibitions imposed by the Greek government during the nighttime and essentially reduced trips during risky hours (Katrakazas et al. 2021).\u003c/p\u003e \u003cp\u003eThe three aforementioned variables are evidently not directly related to the pandemic. Nevertheless, four COVID-19-related variables were found to impact HA and HB events. New COVID-19 cases in Greece were found to prevail compared to other COVID-19-related variables in terms of HA events. On contrary to HA, COVID-19 Reproduction Rate was found to influence HB events the most. The most influential pandemic-related factors for HA and HB events in Greece were COVID-19 Reproduction Rate, Stringency Index, and New COVID-19 Fatalities and Cases. This is in line with existing literature. For example, the studies of (Dong et al. 2022; Lee et al. 2020; Vanlaar et al. 2021) found that COVID-19 restrictions negatively affected risky driving behaviors such as speeding, and distracted driving.\u003c/p\u003e \u003cp\u003eWith regards to traffic exposure during 2020, it can be concluded from Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e that driving requests were significantly decreased during both lockdowns compared to the baseline of no restrictions. The greatest reduction was observed in the first lockdown compared to the second. This means that the traffic volume during the 1st lockdown was lower than in the other conditions (i.e., during the 2nd lockdown, and without restrictions). Hence, with fewer vehicles ahead, the drivers could accelerate more easily and this can be revealed in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (a), where the upper quartile was higher than in other conditions. Additionally, in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (b) for trips with harsh accelerations occurrence, the median was higher during the 1st lockdown than the other conditions (i.e., during 2nd lockdown, and without restrictions) meaning that the HA events were more frequent. This finding can be related partly to speeding, an increase was revealed in the spatial extent of speeding, and in the level of speeding as well as statistically significant differences in speeding before and after the COVID-19 outbreak (Lee et al. 2020). With regards to the 2nd lockdown, for trips with harsh accelerations, the median was higher compared to conditions without restrictions, as a result of the decreased traffic volume but not at the same magnitude as the 1st lockdown, in which the traffic volume was much lower.\u003c/p\u003e \u003cp\u003eWith regards to HB events, in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e (a), again, the upper quartile is greater during the 1st lockdown than other conditions (i.e., during 2nd lockdown, and without restrictions) and combining Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e (b), the median is higher for trips with HB occurrence and this implies that the HB events were more frequent. This finding is also consistent with the literature (Katrakazas et al. 2020). This can be explained, for instance, as the traffic volume during the 1st lockdown was lower than the other conditions and hence with fewer vehicles ahead, the drivers could maintain higher speeds, as stated in (Katrakazas et al. 2020). With higher speeds, the drivers were more probable to be involved in a harsh braking event with potential traffic obstacles ahead (i.e., pedestrians, bikes, scooters and traffic control signs or signals), especially during the lockdowns that the active transport was increased (Linares-Rend\u0026oacute;n and Garrido-Cumbrera 2021). With regards to the 2nd lockdown following the same logic as HA, for trips with harsh brakings, the median was higher compared to no restrictions as a result of the decreased traffic volume but not the same magnitude as the 1st in which the traffic volume was lower.\u003c/p\u003e \u003cp\u003eThe results of the exploratory analysis by XGBoost indicate a correlation of COVID-19 metrics and restrictive measures with harsh brakings and accelerations. This correlation could be explained as COVID-19 metrics and restriction measures affected commuters by leading them to stay at home. Consequently, the stay at home restrictions led to a decreased traffic volume, this can be validated by Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, and thus the traffic volumes affected directly driving behavior. This phenomenon can substantiate why the driving requests are a more important factor in the analysis than COVID-19-related variables.\u003c/p\u003e \u003cp\u003eNevertheless, this work is not without shortcomings, and therefore, future research could focus on covering the remaining gaps that this work did not cover. Initially, future studies could concentrate on more sophisticated models, such as deep neural networks, e.g., Convolutional neural networks (CNNs) or Artificial Neural Networks (ANNs), which probably can accomplish lower errors and give more insights into driving behavior variables. In addition, more variables with regards to driving behavior, i.e., speeding, speeding duration, and speed, could be exploited using the same method in order to give in the same context results. These variables were tested but they led to models with large errors and therefore, were not included in this work. Finally, additional data with geolocation and road type information could also enhance the current methodology, leading to spatial analyses of the examined variables.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThe present paper aims at identifying the most important factors that influenced driving behavior during 2020, a year that behavior was heavily influenced by the COVID-19 pandemic. In order to accomplish this study, naturalistic driving data for a 12-month timeframe were exploited and analyzed. The examined driving behavior variables were HA and HB events per 100km concerning the time period before, during and after the imposition of lockdown measures in Greece in 2020, the first year of the COVID-19 pandemic. The naturalistic driving data were extracted by a specially developed smartphone application and were transmitted to a back-end telematic platform from OSeven Telematics. The top three variables that influenced the most HA and HB events were distance, mobile use/ driving time, and driving requests (as requested in Apple Maps). Focusing on the COVID-19-related variables, this study identified that the most significant factors in the entire 2020 were new COVID-19 cases, new COVID-19 fatalities, COVID-19 reproduction rate as well as stringency index. The results of the exploratory analysis by XGBoost indicate a correlation of COVID-19 metrics and restriction measures with harsh brakings and accelerations. As mentioned previously, COVID-19 restrictions affected commuters to stay at home, and subsequently with lower traffic volumes; driving behavior was affected. Furthermore, for all the investigated three conditions, i.e., no restrictions, 1st lockdown, and 2nd lockdown, different HA and HB event patterns were revealed. HA and HB events were more frequent and with higher values range during the 1st and then 2nd lockdown compared to non-restrictive conditions for trips with harsh events occurrence due to their correlation with driving exposure measurements (i.e., Apple driving requests). Moreover, it was found that HA and HB events were also affected by risky nighttime hours, indicating that there was a change in events during nighttime driving (00:00\u0026ndash;05:00) due to the lighting conditions themselves as well as probably affected due to the prohibitions imposed by the Greek government during the nighttime and essentially reduced trips during risky hours.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthical Approval\u003c/p\u003e\n\u003cp\u003eThe ethics guidelines have been approved by the director of the Department of Transportation Planning and Engineering of the School of Civil Engineering at the National Technical University of Athens. The authors declare that the current study was conducted according to the ethical principles of the Declaration of Helsinki since no one was harmed physically or mentally affected during the driving measurements, and the drivers participated voluntarily.\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eOn behalf of all authors, the corresponding author states that there is no conflict of interest.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAuthors\u0026apos; contributions\u003c/p\u003e\n\u003cp\u003eMarios Sekadakis: Conceptualization, Methodology, Writing \u0026ndash; original draft, Analysis, Data curation. Christos Katrakazas: Conceptualization, Methodology, Writing \u0026ndash; review \u0026amp; editing. Eva Michelaraki: Writing \u0026ndash; original draft, review \u0026amp; editing. Apostolos Ziakopoulos: Analysis, Data curation. George Yannis: Supervision, Resources. All authors reviewed the manuscript.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials\u003c/p\u003e\n\u003cp\u003eThe datasets analysed during the current study are not publicly available due to are confidential in nature and may only be provided with restrictions (e.g. anonymized data) by OSeven Telematics, London, UK (www.oseven.io).\u003c/p\u003e\n\u003cp\u003eAcknowledgment\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank OSeven Telematics, London, UK (oseven.io) for providing all the necessary naturalistic driving data that assisted in accomplishing this study.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003e\u003cspan\u003eAletta, F., Brinchi, S., Carrese, S., Gemma, A., Guattari, C., Mannini, L., and Patella, S. M. (2020). \u0026ldquo;Analysing urban traffic volumes and mapping noise emissions in Rome (Italy) in the context of containment measures for the COVID-19 disease.\u0026rdquo; \u003cem\u003eNoise Mapping\u003c/em\u003e, 7(1), 114\u0026ndash;122.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eApple. (2020). \u0026ldquo;COVID‑19 - Mobility Trends Reports - Apple [WWW Document].\u0026rdquo; \u003cem\u003eAccessed 11/06/20\u003c/em\u003e, \u0026lt;https://www.apple.com/covid19/mobility\u0026gt; (Jun. 11, 2020).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBucsky, P. (2020). \u0026ldquo;Modal share changes due to COVID-19 : The case of Budapest.\u0026rdquo; \u003cem\u003eTransportation Research Interdisciplinary Perspectives\u003c/em\u003e, The Author, 8, 100141.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCarter, D. (2020). \u003cem\u003eEffects of COVID-19 Shutdown on Crashes and Travel in NC\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDong, X., Xie, K., and Yang, H. (2022). \u0026ldquo;How did COVID-19 impact driving behaviors and crash Severity? A multigroup structural equation modeling.\u0026rdquo; \u003cem\u003eAccident Analysis and Prevention\u003c/em\u003e, Elsevier Ltd, 172(December 2021), 106687.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003ede Haas, M., Faber, R., and Hamersma, M. (2020). \u0026ldquo;How COVID-19 and the Dutch \u0026lsquo;intelligent lockdown\u0026rsquo; change activities, work and travel behaviour: Evidence from longitudinal data in the Netherlands.\u0026rdquo; \u003cem\u003eTransportation Research Interdisciplinary Perspectives\u003c/em\u003e, Elsevier Ltd, 6, 100150.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHale, T., Anania, J., Angrist, N., Boby, T., Cameron-Blake, E., Folco, M. Di, Ellen, L., Goldszmidt, R., Hallas, L., KIra, B., Luciano, M., Majumadar, S., Nagesh, R., Petherick, A., Phillips, T., Tatlow, H., Webster, S., Wood, A., and Zhang, Y. (2021). \u003cem\u003eVariation in government responses to COVID-19, Version 12.0\u003c/em\u003e. \u003cem\u003eBSG Working Paper Series\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHale, T., Angrist, N., Cameron-Blake, E., Hallas, L., Kira, B., Majumdar, S., Petherick, A., Phillips, T., Tatlow, H., and Webster, S. (2020). \u0026ldquo;Variation in government responses to COVID-19.\u0026rdquo; \u003cem\u003eBSG Working Paper Series. Blavatnik School of Government. University of Oxford\u003c/em\u003e, Version 8.0.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHuang, Y., and Meng, S. (2019). \u0026ldquo;Automobile insurance classification ratemaking based on telematics driving data.\u0026rdquo; \u003cem\u003eDecision Support Systems\u003c/em\u003e, Elsevier, 127(January 2019), 113156.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKatrakazas, C., Michelaraki, E., Sekadakis, M., and Yannis, G. (2020). \u0026ldquo;A descriptive analysis of the effect of the COVID-19 pandemic on driving behavior and road safety.\u0026rdquo; \u003cem\u003eTransportation Research Interdisciplinary Perspectives\u003c/em\u003e, Elsevier Ltd, 7, 100186.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKatrakazas, C., Michelaraki, E., Sekadakis, M., Ziakopoulos, A., Kontaxi, A., and Yannis, G. (2021). \u0026ldquo;Identifying the impact of the COVID-19 pandemic on driving behavior using naturalistic driving data and time series forecasting.\u0026rdquo; \u003cem\u003eJournal of Safety Research\u003c/em\u003e, National Safety Council and Elsevier Ltd, 78, 189\u0026ndash;202.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKim, K. (2021). \u0026ldquo;Impacts of COVID-19 on transportation: Summary and synthesis of interdisciplinary research.\u0026rdquo; \u003cem\u003eTransportation Research Interdisciplinary Perspectives\u003c/em\u003e, Volume 9(100305).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLee, J., Porr, A., and Miller, H. (2020). \u0026ldquo;Evidence of Increased Vehicle Speeding in Ohio \u0026rsquo; s Major Cities during the COVID-19 Pandemic.\u0026rdquo; 1\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLinares-Rend\u0026oacute;n, F., and Garrido-Cumbrera, M. (2021). \u0026ldquo;Impact of the COVID-19 Pandemic on Urban Mobility: A Systematic Review of the Existing Literature.\u0026rdquo; \u003cem\u003eJournal of Transport \u0026amp; Health\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eMichelaraki, E., Sekadakis, M., Katrakazas, C., Ziakopoulos, A., and Yannis, G. (2021). \u0026ldquo;A four-country comparative overview of the impact of COVID-19 on traffic safety behavior.\u0026rdquo; \u003cem\u003e10th International Congress on Transportation Research\u003c/em\u003e, Rhodes, Greece.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eMurphy, K. P. (2012). \u003cem\u003eMachine learning : a probabilistic perspective\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eNielsen, D. (2016). \u0026ldquo;Tree Boosting With XGBoost: Why Does XGBoost Win \u0026lsquo;Every\u0026rsquo; Machine Learning Competition?\u0026rdquo; Department of Mathematical Sciences.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eOur World in Data. (2020). \u0026ldquo;Coronavirus Pandemic (COVID-19).\u0026rdquo; \u0026lt;https://ourworldindata.org/coronavirus\u0026gt; (Dec. 8, 2020).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSaladi\u0026eacute;, \u0026Ograve;., Bustamante, E., and Guti\u0026eacute;rrez, A. (2020). \u0026ldquo;Transportation Research Interdisciplinary Perspectives COVID-19 lockdown and reduction of traf fi c accidents in Tarragona province, Spain.\u0026rdquo; \u003cem\u003eTransportation Research Interdisciplinary Perspectives\u003c/em\u003e, The Authors, 8, 100218.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSekadakis, M., Katrakazas, C., Michelaraki, E., Kehagia, F., and Yannis, G. (2021). \u0026ldquo;Analysis of the impact of COVID-19 on collisions, fatalities and injuries using time series forecasting: The case of Greece.\u0026rdquo; \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, Elsevier Ltd, 162(July), 106391.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSharifi, A., and Reza Khavarian-Garmsir, A. (2020). \u0026ldquo;The COVID-19 pandemic : Impacts on cities and major lessons for urban planning, design, and management.\u0026rdquo; \u003cem\u003eScience of the Total Environment\u003c/em\u003e, 749, 1\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eShilling, F., and Waetjen, D. (2020). \u0026ldquo;Impact of COVID19 mitigation on numbers and costs of California Traffic Crashes.\u0026rdquo; 2020.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eVanlaar, W. G. M., Woods-Fry, H., Barrett, H., Lyon, C., Brown, S., Wicklund, C., and Robertson, R. D. (2021). \u0026ldquo;The impact of COVID-19 on road safety in Canada and the United States.\u0026rdquo; \u003cem\u003eAccident Analysis and Prevention\u003c/em\u003e, Elsevier Ltd, 160, 106324.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eXGBoost Documentation. (2021). \u0026ldquo;dmlc XGBoost.\u0026rdquo; \u0026lt;https://xgboost.readthedocs.io/en/latest/index.html#xgboost-documentation\u0026gt; (Nov. 26, 2021).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eZhu, N., Zhang, D., Wang, W., Li, X., Yang, B., Song, J., Zhao, X., Huang, B., Shi, W., Lu, R., Niu, P., Zhan, F., Ma, X., Wang, D., Xu, W., Wu, G., Gao, G. F., and Tan, W. (2020). \u0026ldquo;A Novel Coronavirus from Patients with Pneumonia in China, 2019.\u0026rdquo; \u003cem\u003eNew England Journal of Medicine\u003c/em\u003e, 382(8), 727\u0026ndash;733.\u003c/span\u003e\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"COVID-19 pandemic, Driving Behavior, Harsh Brakings, Harsh Accelerations, Feature Analysis, XGBoost","lastPublishedDoi":"10.21203/rs.3.rs-2084342/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-2084342/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis paper tries to identify and investigate the most significant factors that influenced the relationship between COVID-19 pandemic metrics (i.e., COVID-19 cases, fatalities and reproduction rate) and restrictions (i.e., stringency index and lockdown measures) with driving behavior in the entire 2020. To that aim, naturalistic driving data for a 12-month timeframe were exploited and analyzed. The examined driving behavior variables included harsh acceleration and harsh braking events concerning the time period before, during and after the lockdown measures in Greece. The harsh events were extracted using data obtained by a specially developed smartphone application which were transmitted to a back-end telematic platform between the 1st of January and the 31st of December, 2020. Based on the collected data, XGBoost feature analysis algorithms were deployed in order to obtain the most significant factors. Furthermore, a comparison among the first COVID-19 lockdown (i.e., February to May 2020), the second one (i.e., August to November 2020) and the period without COVID-19 restrictions was drawn. COVID-19 new cases and new fatalities were the most significant factors related to COVID-19 metrics impacting driving behavior. Additionally, the correlation between driving behavior with other factors (i.e., distance travelled, mobile use, driving requests, driving during risky hours) was revealed. Furthermore, the differences and similarities of the harsh events between the two lockdown periods were identified. This paper tries to fill this gap in existing literature concerning a feature analysis for the entire 2020 and including the first and second lockdown restrictions of the COVID-19 pandemic in Greece.\u003c/p\u003e","manuscriptTitle":"COVID-19 and driving behavior: Which were the most crucial influencing factors?","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2022-10-25 19:01:34","doi":"10.21203/rs.3.rs-2084342/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9c9a1052-49c0-4db9-8e07-97540916ddc7","owner":[],"postedDate":"October 25th, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2023-10-16T22:23:56+00:00","versionOfRecord":{"articleIdentity":"rs-2084342","link":"https://doi.org/10.1007/s42421-023-00078-7","journal":{"identity":"data-science-for-transportation","isVorOnly":false,"title":"Data Science for Transportation"},"publishedOn":"2023-08-08 21:55:27","publishedOnDateReadable":"August 8th, 2023"},"versionCreatedAt":"2022-10-25 19:01:34","video":"","vorDoi":"10.1007/s42421-023-00078-7","vorDoiUrl":"https://doi.org/10.1007/s42421-023-00078-7","workflowStages":[]},"version":"v1","identity":"rs-2084342","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-2084342","identity":"rs-2084342","version":["v1"]},"buildId":"GqpaHPwrfC8PjnIFayRh5","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}